Missing Types: Overcoming the Typology Dilemma of Lithic Archaeology in Southeast Asia

MISSING TYPES: OVERCOMING THE TYPOLOGY DILEMMA OF LITHIC ARCHAEOLOGY IN SOUTHEAST ASIA.

Miriam Noël Haidle1 and Alfred F. Pawlik2

1University of Tübingen, Germany 2Archaeological Studies Program, University of the Philippines, Diliman, Quezon City Emails: [email protected]; [email protected]

In Europe as well as in the Americas, typology of at least Susan Bulmer (Auckland, New Zealand) mooted artefacts some key forms was and still is the basis of relative from New Guinea which have long been ignored. Stone chronology. For Southeast Asian prehistory, attempts to tool assemblages from five excavations in the Central classify lithic assemblages morphologically and Highlands of Papua New Guinea were restudied: four technologically in order to fit them into established stone rockshelters in and near the Wahgi Valley, and one open- tool typologies from other parts of the world have not air site, a natural swamp that was first cultivated at around proved to be very useful. Up to now, the formation of a 10,000 BP. Bulmer focused mainly on Pleistocene axes specific regional typology system has failed. Session 1C and axe-like tools and compared the evidence of the of the 18th Congress of the Indo-Pacific Prehistory Highlands with two other Pleistocene sites, Bobongara Association at Manila (Fig. 1) wanted to explore the and Kosipe, the former found on the former coastline and “missing types”-problem and adjacent questions. Is the the latter found high on the edge of the upper mountain typological approach not appropriate to the special forest. The types defined are based on empirical attributes characteristics and circumstances of lithic industries in such as size, shape, the position and nature of their Southeast Asia? Why do we know of only a few formal working edges, and the wear they exhibit. A provisional “Asian” types and how did the availability and acquisition chronology of the stone tools has been suggested. of raw material influence lithic technology? Are morphological features of lithic artefacts significant enough at all to establish an acceptable chronology system of lithic periods? How can morphological features of lithic artefacts be linked to geostratigraphy in Southeast Asia? Can non-stratified surface finds contribute to Palaeolithic and Neolithic chronologies in that region? And if we dismiss the typological approach, what are the alternatives? Mirroring the state of discussion in lithic analysis in Southeast Asian archaeology, the contributions of the session covered a broad range of subjects. With his talk titled “River basin archaeology” Israel B. Cabanilla (University of the Philippines) reviewed aspects of early Philippine prehistory and site formation. Palaeolithic sites in the Philippines seemingly date back to 400-500,000 years. While most of the investigated Palaeolithic sites are situated in Northern Luzon and on Palawan Island, Fig. 1: Contributors and participants in Session 1 C “Missing Cabanilla focused in his talk on the river basins of the Types” at the 18th IPPA Conference in Manila: Sue Bulmer, Manila area. In his examination of the vast collection of Helen Selimiotis, Claire Gaillard, Eusebio Dizon, Miriam H. Otley Beyer (1947), a pioneer in Philippine Haidle, Johan Kamminga, Tessa Boer-Mah, Elise Patol- archaeology, and of various surveys of the National Edoumba, Ben Marwick, Nishimura Masanari, Jean-Michel Chazine, Israel Cabanilla, Alfred Pawlik (not in the photo: Museum conducted since the 1960s by Robert Fox and Sharon Teodosio) others, Cabanilla revealed that a major share of lithic artefacts originate from the Manila region and are South East Asian core tool industries were in the associated with the tributaries of Manila Bay and Laguna centre of the paper of Sharon F. Teodosio (University of de Bay like Marilao, Pasig and Santa Mesa. the Philippines). Twenty-one open-air localities in In her presentation on “Pleistocene stone tools of New Northern Luzon along the Cagayan River valley Guinea: a new analysis from the Far East of the Far East”, (Ronquillo 1982) formed the basis for her study

IPPA BULLETIN 29, 2009: 2-5

BULLETIN OF THE INDO-PACIFIC PREHISTORY ASSOCIATION 29, 2009

“Reassessment of the Cabalwanian industry”, which has horsehoofs, side- and end-choppers and a new type called been thought to be of Palaeolithic age because of the ‘flat-iron’. Classically, these tools were interpreted as part crude appearance of some of the tools and their of a hunter-gatherer toolkit, yet the re-examination has association with a Middle-Pleistocene megafauna. This indicated that the site was primarily a pebble adze so-called Cabalwanian industry (Koenigswald 1958) workshop on the river terrace where the availability of consists mainly of flakes and a few large core tools made waterworn stones made the location ideal as a raw of chert and andesite cobbles. Within the category of material source. Kamminga saw no evidence that the heavy-duty core tools, Fox (1978) identified six different adzes were made by hunter-gatherers and instead tool types, which Teodosio, however, has found hard to discussed an association with early horticulturalists since assess. On the basis of the material from the recently the tool type is functionally equivalent to the general discovered open-air site of Arubo in Central Luzon, category of East Asian and Pacific ground stone adzes which includes a proto-handaxe and quite sophisticated that succeeded them. core technology, she has argued for a technological and Three papers of the session dealt with seemingly functional approach to be much more promising than a conservative lithic technology from late hunter-gatherer to typological-chronological one, for which – up to now – early Neolithic groups. In his talk about “Recent adequate sites with well observed context are lacking. discoveries and results from Kalimantan Timur, With the diachronic approach of his talk on “Lithic Indonesian Borneo (2003- 2005)”. Jean-Michel Chazine assemblage problems in the Palaeolithic and (Centre de Recherche et de Documentation sur l'Océanie) Epipalaeolithic Age of Northern Vietnam and Guangxi- presented a series of cave sites in three different altitudes Guangdong” Nishimura Masanari (Center for Vietnamese of the Marang conical karst in East Kalimantan. The and Inter-cultural Studies, Ha Noi National University) group at the lowest altitude level revealed occupation sites linked two chronological foci in his contribution to this with stratigraphies from at least the end of the Pleistocene session. He reviewed the development of lithic industries through the Holocene. The stone tool workshops studied in North Vietnam from Lower Pleistocene to End- by Julien Espagne exhibited a Pleistocene knapping Pleistocene and compared the Lower Palaeolithic Son- technology throughout the Epipalaeolithic until the early Vian assemblages from Lang Vac to the Bose industry Neolithic sequences with a classical flake technology and from South China. Stratified sites with flake and core tool the typical Kutai flakes. On large flakes the bulbar region industries characterized by a considerable amount of was transversely detached with a secondary blow. The Hoabinhian tools were also emphasized. resulting Kutai flakes were then used either as blanks for An approach to the characterization of Hoabinhian tools or as micro-cores as a basis for a tertiary debitage assemblages themselves was given by Ben Marwick stage. (Australian National University) in his talk “Beyond Helen Selimiotis (Archaeology and Anthropology, typologies: The reduction thesis and its implications for Australian National University) studied the “Core lithic assemblages in Southeast Asia” (Marwick 2008). technology at Bui Ceri Uato, East Timor”, a site with a Two different analytical systems have to be distinguished: cultural sequence beginning from more than 26,000 years a) essentialism, looking for discrete categories (typology), BP. This period covers the replacement of endemic fauna and b) nominalism, stating that there are no discrete types (primarily murids) in aceramic lower deposits with exotic but only tendencies which can be abstracted (e.g. fauna (such as pig, dog, Capra/Ovis, Bos, Macaca) technological reduction approach). The reduction thesis is seemingly associated with pottery in upper levels. The based on the assumption that artefacts are continuously faunal change has been interpreted as a transition from a reworked during their life cycles and that consequently hunter-gatherer subsistence economy to animal observed shape variation is only a variation in the husbandry, presumably some sort of agricultural reduction pattern. Taking Hoabinhian assemblages as an practices, and the establishment of village communities. example, Marwick showed how reduction can be At Bui Ceri Uato, Selimiotis examined the apparent measured and how the differential distribution of continuity in East Timor lithic traditions described by Ian reduction stages can be used for behavioural modelling Glover (1986). She applied a technological analysis that and the measurement of change differences. focuses on the extent/length of core reduction and thermal The chronological and subsistence attribution of fractures of cores as an indicator of fire/hearth use. Along Hoabinhian complexes was discussed by Johan with sedimentation rates and discard rates of flaked stone Kamminga (National Heritage Consultants, Canberra). He artefacts, these factors have been used to describe the presented a reassessment of the “Hoabinhian stone intensity of occupation and landscape use. technology at Sai Yok, Kanchanaburi Province, Central In her talk about “A typo-technological definition of Thailand”, the first large-scale controlled excavation of a Tabonian industries”, Elise Patol-Edoumba (Natural prehistoric site in Thailand realized by a Thai-Danish History Museum, La Rochelle, France) summarized the research group in the early 1960s. Classified by van analyses of six lithic assemblages from five sites on Heekeren (1962) in the traditional Eurocentric framework Palawan Island (Tabon Cave, Guri Cave and the Duyong established by Madeleine Colani (1927), the stone rockshelter) and in the limestone formation of Penablanca artefacts provided an account of a classic Hoabinhian in Cagayan, Northern Luzon (Laurente Cave and Musang assemblage with sumatraliths, short-axes, picks, Cave) dating between 12,000 and 4000 BP. The Tabonian

3 HAIDLE AND PAWLIK: MISSING TYPES

industry, with chert as its main raw material, can be typology using the functional interpretation of lithic characterized as a flake industry with around 14 % of artefacts. The identification of functional types is possible tools and a high percentage of blades over 20%. No even in those assemblages where morpho- types are few morpho-types could be identified, but a detailed analysis or missing. In addition to techno-types, the functional of the chaînes opératoires showed nine recurrent techno- types would help to reassess such seemingly simple or types. Although the overall setting of the Tabonian “primitive” lithic assemblages. A more realistic industry remains the same bridging the late hunter- characterization of stone tools should consider their actual gatherer to early Neolithic transformation, chronological uses and purposes, and would consequently be able to and geographical variations between Palawan and give more detailed information about the prehistoric life, Northern Luzon have been identified. technology, and subsistence of the people who produced Later chronological stages, namely the Neolithic and and used them. the early Metal Ages are touched on by two papers in this The present situation of lithic archaeology in session, and both looking into ground stone artefacts. The Southeast Asia might not appear as promising as in other first, by Tessa Boer-Mah (University of Sydney), follows regions of the world. The number of lithic archaeologists a similar approach as Ben Marwick in questioning the is still few. The detection of well-preserved open-air sites traditional classification of “Ground stone adzes from is difficult in the humid tropical areas. They are either Northeast Thailand” within a cultural-historical located on river terraces and prone to all problems of framework of human migration and the spread of the identification, or the artefacts are found in the waste heap Neolithic (Boer-Mah 2008). Heine-Geldern (1932) and of modern construction work like the proto-handaxe of Duff (1970) proposed typologies and identified ‘cultures’ Arubo in Northern Luzon. Few excavations, be it in caves on the basis of inter-regional comparisons. In her study or open-air sites, have been realized with techniques focusing on an adze assemblage from a single site, Ban appropriate to allow the complete range of modern lithic Non Wat, Boer-Mah has demonstrated that factors like analyses. In comparison with European (typological) data reduction, raw material and the systemic context of the the Southeast Asian data set is small and not very tool have a significant impact on the form of each adze. promising. However, typology based on morphological These adze forms have not necessarily been static. The classifications is only one – with limited explanatory morphology may have changed throughout its life-history: potential – of several approaches to lithic assemblages. four reduction sequences have been identified in the lithic First of all, the raw material economy of a site can be assemblage from Ban Non Wat. Therefore, the variation explored by searching for resource locations and raw within adze assemblages has to be seen as a technological material procurement, by examining the exploitation of outcome rather than a typological set. the original blanks and the curation of the tools using, for An attempt to characterize assemblages from a example, the reduction intensity approach. Different different perspective was made by Miriam Noël Haidle technological concepts or ‘techno-types’ can be traced by (University of Tübingen) in her talk on “Bifaces, analysing the chaînes opératoires which were used to shouldered adzes, and chert prisms: the chronological work on stone raw material. They are supported by potential of stone tool assemblages in Cambodian functional analyses which are able to empirically younger prehistory”. She refrained from focussing on one distinguish functional parts of a tool. Microwear analyses lithic artefact class – be it typological, technological or can extend our knowledge by giving information on functional – because of the great diversity in several activities done with the lithic artefacts resulting in dimensions of the lithic assemblages from the late different classification systems of functional types. Neolithic to early Metal Age complex of the Mimotien Technological analyses of the production debris, blanks, circular earthwork sites. At least eight different stone raw semi-finished products and dumped exhausted pieces, materials have been identified as well as 13 broad artefact together with spatial analyses of the different activities at classes, seven modes of production, and nine probable the site will yield a more detailed picture about the modes of use. The presumable on-site application of a organisation of prehistoric way of life. production mode was indicated by the presence of More or less basic to all these approaches are different production stages and waste. The combination of archaeological assemblages which have not been selected the different variables has yielded a specific pattern that for apparent morpho-types but are accompanied by as can be used to detect differences between sites, regions much additional information as possible. Lithic studies and periods. can tell a lot about technical knowledge and skills, In the last talk of the session Alfred Pawlik environmental perception, economic organisation, and (University of the Philippines) viewed the Southeast cultural development, like the conservatism in lithic Asian lithic typology dilemma from “Typology, technology at the transition from the upper Palaeolithic to technology and function: A use-wear analyst’s the early Neolithic. It would be desirable if the morpho- perspective” and proposed a functional approach. Based types could serve as clear-cut indicators in a on the examination of wear traces on artefact surfaces by straightforward chronological classification system, using optical light microscopy, chemical analysis and which, in the Palaeolithic, concentrated on clearly SEM, micro-wear analysis can not only yield information distinguishable stages and, in later periods, fitted to about the use of artefacts, but result in a new form of subdivisions by other artefact groups like ceramics or

4 BULLETIN OF THE INDO-PACIFIC PREHISTORY ASSOCIATION 29, 2009

subsistence patterns. Yet, like world and cultural history, Fox, Robert B. 1978. The Philippine Paleolithic. In: Fumiko the lithic element is not that simple and its technological Ikawa-Smith (ed.), Early Paleolithic in South and East development does not follow a straight line. A Asia, pp. 59-85. The Hague: Mouton. chronological classification needs to be based on faunal, Glover, Ian C. 1986. Archaeology in Eastern Timor, 1966-67. stratigraphic, radiocarbon or other independent evidence, Canberra: Department of Prehistory, Research School of allowing a characterization of the sites as well as regional Pacific Studies, Australian National University. and subsequently defined chronological groups by several Heekeren, Hendrik R. van 1962. A brief survey of the Sai-Yok different typologies: morpho-types, techno-types, excavations, 1961-1962 season of the Thai-Danish functional types, raw material management patterns etc. prehistoric expedition. Journal of the Siam Society 50: 15- 18. Despite all difficulties, the participants of Session 1C concluded that Southeast Asian lithic assemblages Heine-Geldern, Robert von 1932. Urheimat und früheste deserve appropriate analyses according to their individual Wanderungen der Austronesier. Anthropos 27: 543-619 potentials, as do lithic assemblages in the rest of the Koenigswald, G.H. Ralph von 1958: Preliminary report on a world. newly-discovered Stone Age culture from northern Luzon, Philippine Islands. Asian Perspectives II (2): 69- 71. REFERENCES Marwick, B. 2008. Beyond typologies: The reduction thesis and Beyer, H. Otley 1947. Outline Review of Philippine its implications for lithic assemblages in Southeast Asia. Archaeology by Islands and Provinces. Philippine Journal IPPA Bulletin 28: 108-116. of Science, Vol. 77, no. 3-4. Manila. Ronquillo, Wilfredo P. 1982. Cagayan Valley Archaeological Boer-Mah, T. 2008. Reduction, raw materials and form: ground Project. Manuscript of the National Museum of the stone adzes from Ban Non Wat, Northeast Thailand. IPPA Philippines, Manila. Bulletin 28: 44-51. Colani, Madeleine 1927. L’âge de la pierre dans la provnce de Hoa Binh, Tonkin. Mémoires du Service Géologique du l’Indochine XIV, fasc. 1, Hanoi. Duff, Roger 1970. Stone Adzes of Southeast Asia: An Illustrated Typology. Christchurch: Canterbury Museum.